We investigated whether carboxylate exudation by chickpea (Cicer arietinum L.) was affected by soil bulk density and if this effect was local or systemic. We hypothesised that concentrations of carboxylates would increase with distance from the root apex due to continuous and constitutive accumulation of carboxylates, and that exudate accumulation would be greater in a compacted soil than in a loose soil. Plants were grown in split-root or single cylinders containing loose (1400 kg m(-3)) or compacted (1800 kg m(-3)) soil. Rhizosphere carboxylate concentrations were measured of whole root systems as well as of sections along the root. The root diameter was greatest of plants grown in the compacted soil; however, root diameters were the same for both root halves in the split-root design, whether they grew in loose soil or in compacted soil. Similarly, carboxylate concentrations tended to be lower for the whole root system in the compacted soil, but were the same for both root halves in the split-root design, irrespective of whether the roots were in loose soil or in compacted soil. These results indicate that both root diameter and carboxylate exudation by roots in chickpea is regulated systemically via a signal from the shoot rather than by local signals in the roots. There was no accumulation of carboxylates with increasing distance from the apex, probably because microbial degradation occurred at similar rates as carboxylate exudation. Malonate, previously suggested as deterrent to microorganisms, is likely only a selective deterrent.